gt2500 cancer genomes Overall functional impact Slides freely downloadable from LecturesGersteinLaborg amp tweetable via markgerstein See last slide for more info ID: 660953
Download Presentation The PPT/PDF document "Passenger Mutations in" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Passenger Mutations
in >2500 cancer genomes: Overall functional impact
Slides freely downloadable from Lectures.GersteinLab.org & “tweetable” (via @markgerstein) See last slide for more info.
Mark Gerstein
YaleSlide2
Drivers
directly confer a selective growth advantage to the tumor cell.A typical tumor contains 2-8 drivers.identified through signals of positive selection.Existing cohorts of ~100s give enough power to identify
PassengersConceptually, a passenger mutation has no direct or indirect effect on tumor progression.There are 1000s of passengers in a typical cancer genome.
Canonical model of drivers
&
passengers in cancer
[
Vogelstein
Science
2013.
339:1546]Slide3
Conceptual extension
of the canonical model of drivers & passengersSlide4
S: Mutation signature
inferred
M: Mutation spectrum
observed
[T.
Helleday,
S.
Eshtad &
S.
Nik-Zainal,
Nature Reviews Genetics
(
’
14
), L.
Alexandrov
et al., Nature (‘13) ]
Mutational processes carry context-specific signatures
A[C>T]G
C[C>T]G
M = S × W+
εSlide5
PCAWG : most comprehensive resource for
cancer
whole genome analysisProject Goals:To understand role of non-coding regions of cancer genomes in cancer progression.
Union of TCGA-ICGC efforts
Jointly analyzing ~2800 whole genome tumor/normal pairs
> 580 researchers
16 thematic working groups
~30M total somatic SNVs
Adapted from Campbell et. al.
,
bioRxiv
(
’
1
7)Slide6
A case study:
pRCCKidney cancer lifetime risk of 1.6% & the p
apillary type (pRCC) counts for ~10% of all casesTCGA project sequenced 161 pRCC exomes & classified them into subtypesAlso, 35 WGS of TN pairs
[Cancer Genome Atlas Research Network N
Engl
J Med. (‘16) ]Slide7
Passenger mutations in >2500 cancer genomes:
Overall molecular functional impact
Introduction Background: driver-passenger model (w/ conceptual extension) + mutational spectra & signaturesData source: PCAWG comprehensive WGS on >2.5K + focus on 35 pRCC WGSOverall functional impact of variantsFunSeq entropy-weights multiple features to evaluate the functional impact of SNVsInvestigating how the fraction of high-impact (non-strong-driver) SNVs scales & how it relates to survival
Differential burdening from various mutational processes
Diff. burdening of TF
sub-networks results
from spectra & signatures differentially affecting binding motifs
High & low impact mutations assoc. w/ diff. signatures
Number of mutations in
DHSes
assoc. w/ specific chromatin mod. mutation
Functional impact & tumor evolution
Mutational timing & tree topology classifies
pRCC
subtypes
Differences
in functional impact
betw
. early & late passenger mutations (
eg
in TSGs & oncogenes)Slide8
Passenger mutations in >2500 cancer genomes:
Overall molecular functional impact
Introduction Background: driver-passenger model (w/ conceptual extension) + mutational spectra & signaturesData source: PCAWG comprehensive WGS on >2.5K + focus on 35 pRCC WGS
Overall functional impact of variants
FunSeq
entropy-weights multiple features to evaluate the functional impact of SNVs
Investigating how the fraction of high-impact (non-strong-driver) SNVs scales & how it relates to survival
Differential burdening from various mutational processes
Diff. burdening of TF
sub-networks results
from spectra & signatures differentially affecting binding motifs
High & low impact mutations assoc. w/ diff. signatures
Number of mutations in
DHSes
assoc. w/ specific chromatin mod. mutation
Functional impact & tumor evolution
Mutational timing & tree topology classifies
pRCC
subtypes
Differences
in functional impact
betw
. early & late passenger mutations (
eg
in TSGs & oncogenes)Slide9
Funseq: a flexible framework to
determine functional impact & use this to prioritize
variantsAnnotation (tf binding sites open chromatin, ncRNAs) & Chromatin DynamicsConservation(GERP, allele freq.)Mutational impact (motif breaking, Lof) Network (centrality position) [Fu et al., GenomeBiology
('14),
,
Khurana
et al., Science ('13)]
Slide10
FunSeq
.gersteinlab.org
HOT regionSensitive regionPolymorphisms
Genome
Entropy based method for weighting consistently many genomic features
Practical web server
Submission of variants & pre-computed large
d
ata
context from uniformly processing large-scale datasets
[Fu et al.,
GenomeBiology
('14)]
Slide11
Overall functional impact distribution
of
PCAWG mutations
Funseq
molecular functional
impact
of ~30M variants
in >2500 PCAWG samples
Division of PCAWG Lymph-CLL cohort based on average impact of non-driver variants (high v
low
)
[A result of selection?]Slide12
In many PCAWG cohorts,
the
fraction of impactful “passengers” decreases with increase in total mutation burden(A result of selection?)Slide13
Passenger mutations in >2500 cancer genomes:
Overall molecular functional impact
Introduction Background: driver-passenger model (w/ conceptual extension) + mutational spectra & signaturesData source: PCAWG comprehensive WGS on >2.5K + focus on 35 pRCC WGS
Overall functional impact of variants
FunSeq
entropy-weights multiple features to evaluate the functional impact of SNVs
Investigating how the fraction of high-impact (non-strong-driver) SNVs scales & how it relates to survival
Differential burdening from various mutational processes
Diff. burdening of TF
sub-networks results
from spectra & signatures differentially affecting binding motifs
High & low impact mutations assoc. w/ diff. signatures
Number of mutations in
DHSes
assoc. w/ specific chromatin mod. mutation
Functional impact & tumor evolution
Mutational timing & tree topology classifies
pRCC
subtypes
Differences
in functional impact
betw
. early & late passenger mutations (
eg
in TSGs & oncogenes)Slide14
Differential Mutational burdening of TF-subnetworks due to SNVs breaking & creating binding sites
ARNT
EP300
LOSS
GAIN
ETS regulated genes
ETSSlide15
Kidney cancer as an example:
differential
burdening correlates with mutational spectrumc
oding
LoF
: premature stops (N = 525)
noncoding
LoF
motif breaks in
…
HDAC2 (N=675)
EWSR1 (N=514)
SP1 (N=571)
a
ll mutations (N=
923,782
)Slide16
The loadings on PC1 are mostly [C>T]G
Confirmed by higher C>T% in CpGs in the
hypermethylated group (cluster1) A[C>T]G
C[C>T]G
G[C>T]G
T[C>T]G
CpGs
drive inter-patient variation in
pRCC
mutational spectra
[S. Li, B.
Shuch
and M. Gerstein PLOS Genetics (‘17)] Slide17
Signatures burden the
genome disproportionally
We found 1 pRCC has ApoBEC signature, but nothing in a larger ccRCC cohort
Signatures in
pRCC
[S. Li, B.
Shuch
and M. Gerstein PLOS Genetics (‘17)]
pRCC
ccRCC
enrichment
-log(p-value)
3.6
7.2
22
total mutations
50
150
100
pRCC
ccRCC
high impact passenger SNVs
low
impact passenger SNVs
LoF
SNVs (premature stops)Slide18
C
hromatin remodeling defect (“mut”) leads to more mutations in open chromatin (raw number & fraction) in those
pRCC cases w/ the mutation
Key mutation affects mutational landscape which, in turn, affects overall burden in
pRCC
[S. Li, B.
Shuch
and M. Gerstein PLOS Genetics (‘17)] Slide19
Passenger mutations in >2500 cancer genomes:
Overall molecular functional impact
Introduction Background: driver-passenger model (w/ conceptual extension) + mutational spectra & signaturesData source: PCAWG comprehensive WGS on >2.5K + focus on 35 pRCC WGS
Overall functional impact of variants
FunSeq
entropy-weights multiple features to evaluate the functional impact of SNVs
Investigating how the fraction of high-impact (non-strong-driver) SNVs scales & how it relates to survival
Differential burdening from various mutational processes
Diff. burdening of TF
sub-networks results
from spectra & signatures differentially affecting binding motifs
High & low impact mutations assoc. w/ diff. signatures
Number of mutations in
DHSes
assoc. w/ specific chromatin mod. mutation
Functional impact & tumor evolution
Mutational timing & tree topology classifies
pRCC
subtypes
Differences
in functional impact
betw
. early & late passenger mutations (
eg
in TSGs & oncogenes)Slide20
Construct
ing evolutionary trees in pRCC
Infer mutation order (eg early v late) & tree toplogy based on mutation abundance (PhyloWGS, Deshwar et al., 2015)Some key mutations occur in all the clones while others are just in parts
of the tree
DNMT3A
: premature stop
NEAT1
: noncoding
SMARCA4
: missense
MET
: noncoding
ERRFI1
: noncoding
KDM6A
: missense
[S. Li, B. Shuch and M. Gerstein PLOS Genetics (‘17)]
Slide21
[S. Li, B. Shuch and M. Gerstein PLOS Genetics (‘17)]
MutationdistanceGermline0.5Populations(%)Slide22
[S. Li, B. Shuch and M. Gerstein PLOS Genetics (‘17)]
Mutation
distance
Germline
0.5
Populations
(%)Slide23
Tree topology correlates with molecular subtypes
[Li et al., PLOS Genetics (‘17)] Slide24
Sub-clonal architecture
of mutations
in PCAWGAs expected, drivers are enriched in earlier subclones. Overall, no such enrichment among passengers.High impact passengers are slightly enriched among early subclones(weak drivers?)Particularly, passengers in tumor suppressor (in contrast to oncogenes, which require specific mutations). Slide25
Continuous correlation of
functional impact & VAF
Among mutations in driver genes: higher impact mutation Still true after removing all known driver variants from driver genes. (Latent drivers?)Outside driver genes: higher impact mutation (Deleterious passengers?)
Functional Impact (GERP score)
Early vs Late (mean VAF)Slide26
Passenger mutations in >2500 cancer genomes:
Overall molecular functional impact
Introduction Background: driver-passenger model (w/ conceptual extension) + mutational spectra & signaturesData source: PCAWG comprehensive WGS on >2.5K + focus on 35 pRCC WGS
Overall functional impact of variants
FunSeq
entropy-weights multiple features to evaluate the functional impact of SNVs
Investigating how the fraction of high-impact (non-strong-driver) SNVs scales & how it relates to survival
Differential burdening from various mutational processes
Diff. burdening of TF
sub-networks results
from spectra & signatures differentially affecting binding motifs
High & low impact mutations assoc. w/ diff. signatures
Number of mutations in
DHSes
assoc. w/ specific chromatin mod. mutation
Functional impact & tumor evolution
Mutational timing & tree topology classifies
pRCC
subtypes
Differences
in functional impact
betw
. early & late passenger mutations (
eg
in TSGs & oncogenes)Slide27
Passenger mutations in >2500 cancer genomes:
Overall molecular functional impact
Introduction Background: driver-passenger model (w/ conceptual extension) + mutational spectra & signaturesData source: PCAWG comprehensive WGS on >2.5K + focus on 35 pRCC WGSOverall functional impact of variantsFunSeq entropy-weights multiple features to evaluate the functional impact of SNVsInvestigating how the fraction of high-impact (non-strong-driver) SNVs scales & how it relates to survival
Differential burdening from various mutational processes
Diff. burdening of TF
sub-networks results
from spectra & signatures differentially affecting binding motifs
High & low impact mutations assoc. w/ diff. signatures
Number of mutations in
DHSes
assoc. w/ specific chromatin mod. mutation
Functional impact & tumor evolution
Mutational timing & tree topology classifies
pRCC
subtypes
Differences
in functional impact
betw
. early & late passenger mutations (
eg
in TSGs & oncogenes)Slide28
Acknowledgements
Hiring Postdocs, See
JOBS
.gersteinlab.org
PanCancer.info
Functional impact
S
Kumar
,
J
Warrell
, W Meyerson, P
McGillivary
,
L
Salichos
, S Li, A
Fundichely
, E
Khurana
, C Chan, M Nielsen
,
C
Herrman
, A
Harmanci
,
L
Lochovsky,Y
Zhang, X Li,
PCAWG Drivers & Functional Interpretation Group
(leaders: G
Getz,
J
Pedersen,
J
Stuart,
B
Rapheal
,
N
Lopez
Bigas,
D Wheeler), ICGC/TCGA PCAWG
Network
FunSeq
.gersteinlab.orgY Fu
, E Khurana, Z Liu,
S Lou, J Bedford,
XJ Mu, KY Yip
pRCC
S
Li
,
B
ShuchSlide29Slide30
Info about this talk
General PERMISSIONSThis Presentation is copyright Mark Gerstein, Yale University, 2016. Please read permissions statement at gersteinlab.org/misc/permissions.html .Feel free to use slides & images in the talk with PROPER acknowledgement (via citation to relevant papers or link to appropriate site). Paper
references in the talk probably from Papers.GersteinLab.org. PHOTOS & IMAGES For thoughts on the source and permissions of many of the photos and clipped images in this presentation see streams.gerstein.info . In particular, many of the images have particular EXIF tags, such as kwpotppt , that can be easily queried from flickr, viz: flickr.com/photos/mbgmbg/tags/kwpotppt